1. Field of the Invention
The present invention relates to transistors and methods of making transistors in semiconductor devices. More particularly, the present invention relates to transistors formed of multiple materials having differing work functions.
2. State of the Art
Transistor devices are used with semiconductor devices for numerous purposes, and such use is well known. The characteristics of transistor devices are also well known and documented so that further research may improve the transistor devices. For example, in the case of NMOS transistor devices, it is well known that the drive current of an NMOS transistor device will be higher when a high work function gate material is used as opposed to a low work function gate material. The drive current is stronger in a high work function material because the substrate doping can be much lower with the high work function material, resulting in mobility improvement and the improved drive current.
Similar to NMOS transistor devices, access transistor devices used with memory devices, such as DRAM memory, exhibit a higher drive current when a high work function material is used to form the access transistor as compared to when a lower work function material is used. However, the use of a high work function material to form an access transistor in a memory device may lead to off-state leakage across the access transistor. Off-state leakage includes current leakage that occurs when the access transistor is in an “off” state. Typically, off-state leakage includes two types of leakage: sub-threshold leakage between a source and a drain region associated with the access transistor and leakage between the drain and the substrate of an access device. The leakage from the drain to the substrate may include both junction leakage and gate-induced drain leakage. Junction leakage may include Schokley-Read-Hall type junction leakage and is undesirable. Gate-induced drain leakage (GIDL) is also undesirable.
Recessed access devices (RADs) used as access transistors in memory devices are especially susceptible to gate-induced drain leakage when in an “off” state. The gate-induced drain leakage of a RAD structure dominates the off-state leakage that occurs with such devices. Thus the refresh rate of a RAD structure, and a memory device employing RAD structures, may be dependent upon the amount of gate-induced drain leakage in the RAD device.
Therefore, it is desirable to reduce the amount of gate-induced drain leakage in a RAD structure. It is also desirable to reduce the amount of gate-induced drain leakage while controlling or reducing the amount of other leakages present in the RAD structure or access transistor.